.stringdict plural forms incorrect

[dwaynebailey]

Describe the issue

Comparing Android to iOS plural forms for Afrikaans. They are different while both claiming to be using CLDR. I suspect this is only a visual anomaly but it is enough to confuse translators.

Android:

• https://hosted.weblate.org/translate/organicmaps/android/af/?checksum=e8e3cb8cdbc14c55
• Plural form: n != 1
• Results in:

Android
integer -> index -> cldr form
0 -> 0 -> other
1 -> 1 -> one
2 -> 0 -> other
3 -> 0 -> other
4 -> 0 -> other
5 -> 0 -> other
10 -> 0 -> other

iOS:

• https://hosted.weblate.org/translate/organicmaps/wip-ios-plurals-please-dont-use/af/?checksum=efcf782307b0b6f3
• Plural form: n==0 ? 0 : n!=1 ? 1 : 2
• I am aware of the iOS zero injection for stringdict, so that is ignored
• Results on:

iOS
integer -> index -> cldr form
0 -> 0 -> zero
1 -> 2 -> other
2 -> 1 -> one
3 -> 1 -> one
4 -> 1 -> one
5 -> 1 -> one
10 -> 1 -> one

I've traced the problem here:
https://github.com/WeblateOrg/weblate/blob/stable/weblate/lang/data.py#L235
but I'm not sure:

1. How these where derived
2. If they are merely informational and for providing the CLDR examples, or
3. If they are actually what Android uses

I already tried

• I've read and searched the documentation.
• I've searched for similar filed issues in this repository.

Steps to reproduce the behavior

1. Ensure you have iOS stringdict and Android projects for similar strings
2. Enable Afrikaans (af) which has two forms (one, other)
3. Compare the suggested numeric help between Android and iOS one and other are swapped

Expected behavior

Honour the CLDR forms in iOS plurals https://www.unicode.org/cldr/charts/42/supplemental/language_plural_rules.html#af

The correct formula would be: n==0 ? 0 : n==1 ? 1 : 2

Screenshots

No response

Exception traceback

No response

How do you run Weblate?

weblate.org service

Weblate versions

No response

Weblate deploy checks

No response

Additional context

No response

[nijel]

See #6368

[dwaynebailey]

@nijel that seems to be about zero, this is about one and other being swapped

[dwaynebailey]

Something similar occurs for n>1 equations, note the actual output is correct but the numeric examples for one and other are swapped:

Persian (fa) Android: https://hosted.weblate.org/translate/organicmaps/android/fa/?checksum=e8e3cb8cdbc14c55
Persian (fa) iOS: https://hosted.weblate.org/translate/organicmaps/wip-ios-plurals-please-dont-use/fa/?checksum=efcf782307b0b6f3

[nijel]

Sorry, I didn't understand this from the initial post.

These are only examples shown in Weblate, it should not affect anything else. But you're right that most of (if not all) the formulas have messed up ordering:

weblate/weblate/lang/data.py

Lines 232 to 262 in f006ac8

	FORMULA_WITH_ZERO = {
	"0": "n==0 ? 0 : 1",
	"n > 1": "n==0 ? 0 : n>1 ? 1 : 2",
	"n != 1": "n==0 ? 0 : n!=1 ? 1 : 2",
	"(n == 0 || n == 1) ? 0 : ((n >= 2 && n <= 10) ? 1 : 2)": "n == 0 ? 0 : n == 1 ? 1 : ((n >= 2 && n <= 10) ? 2 : 3)",
	"n == 1 ? 0 : n == 2 ? 1 : 2": "n == 0 ? 0 : n == 1 ? 1 : n == 2 ? 2 : 3",
	"n==1 || n%10==1 ? 0 : 1": "n == 0 ? 0 : n==1 || n%10==1 ? 1 : 2",
	"n==1 ? 0 : n==2 ? 1 : 2": "n==0 ? 0 : n==1 ? 1 : n==2 ? 2 : 3",
	"(n==1 || n==11) ? 0 : (n==2 || n==12) ? 1 : (n > 2 && n < 20) ? 2 : 3": "n==0 ? 0 : (n==1 || n==11) ? 1 : (n==2 || n==12) ? 2 : (n > 2 && n < 20) ? 3 : 4",
	"n != 1 && n != 2 && n != 3 && (n % 10 == 4 || n % 10 == 6 || n % 10 == 9)": "n==0 ? 0 : n != 1 && n != 2 && n != 3 && (n % 10 == 4 || n % 10 == 6 || n % 10 == 9) ? 1: 2",
	"n==1 ? 0 : (n==0 || (n%100 > 0 && n%100 < 20)) ? 1 : 2": "n==0 ? 0 : n==1 ? 1 : (n==0 || (n%100 > 0 && n%100 < 20)) ? 2 : 3",
	"(n == 1) ? 0 : ((n == 0 || n % 100 >= 2 && n % 100 <= 19) ? 1 : 2)": "n==0 ? 0 : (n == 1) ? 1 : ((n == 0 || n % 100 >= 2 && n % 100 <= 19) ? 2 : 3)",
	"n==1 ? 0 : n%10>=2 && n%10<=4 && (n%100<10 || n%100>=20) ? 1 : 2": "n==0 ? 0 : n==1 ? 1 : n%10>=2 && n%10<=4 && (n%100<10 || n%100>=20) ? 2 : 3",
	"(n==1) ? 0 : (n>=2 && n<=4) ? 1 : 2": "n==0 ? 0 : (n==1) ? 1 : (n>=2 && n<=4) ? 2 : 3",
	"(n == 1) ? 0 : ((n == 2) ? 1 : 2)": "n==0 ? 0 : (n == 1) ? 1 : ((n == 2) ? 2 : 3)",
	"n==1 ? 0 : n==0 || ( n%100>1 && n%100<11) ? 1 : (n%100>10 && n%100<20 ) ? 2 : 3": "n==0 ? 0 : n==1 ? 1 : ( n%100>1 && n%100<11) ? 2 : (n%100>10 && n%100<20 ) ? 3 : 4",
	"n%10==1 && n%100!=11 ? 0 : n%10>=2 && n%10<=4 && (n%100<10 || n%100>=20) ? 1 : 2": "n==0 ? 0 : n%10==1 && n%100!=11 ? 1 : n%10>=2 && n%10<=4 && (n%100<10 || n%100>=20) ? 2 : 3",
	"(n == 1) ? 0 : ((n == 2) ? 1 : ((n > 10 && n % 10 == 0) ? 2 : 3))": "n==0 ? 0 : (n == 1) ? 1 : ((n == 2) ? 2 : ((n > 10 && n % 10 == 0) ? 3 : 4))",
	"n==1 ? 0 : n==2 ? 1 : (n>2 && n<7) ? 2 :(n>6 && n<11) ? 3 : 4": "n==0 ? 0 : n==1 ? 1 : n==2 ? 2 : (n>2 && n<7) ? 3 :(n>6 && n<11) ? 4 : 5",
	"(n % 100 == 1) ? 0 : ((n % 100 == 2) ? 1 : ((n % 100 == 3 || n % 100 == 4) ? 2 : 3))": "n==0 ? 0 : (n % 100 == 1) ? 1 : ((n % 100 == 2) ? 2 : ((n % 100 == 3 || n % 100 == 4) ? 3 : 4))",
	"n%100==1 ? 0 : n%100==2 ? 1 : n%100==3 || n%100==4 ? 2 : 3": "n==0 ? 0 : n%100==1 ? 1 : n%100==2 ? 2 : n%100==3 || n%100==4 ? 3 : 4",
	"n % 10 != 1 || n % 100 == 11": "n==0 ? 0 : n % 10 != 1 || n % 100 == 11 ? 1 :2",
	"(n % 10 == 1 && (n % 100 < 11 || n % 100 > 19)) ? 0 : ((n % 10 >= 2 && n % 10 <= 9 && (n % 100 < 11 || n % 100 > 19)) ? 1 : 2)": "n==0 ? 0 : (n % 10 == 1 && (n % 100 < 11 || n % 100 > 19)) ? 1 : ((n % 10 >= 2 && n % 10 <= 9 && (n % 100 < 11 || n % 100 > 19)) ? 2 : 3)",
	"n >= 2 && (n < 11 || n > 99)": "n==0 ? 0 : n >= 2 && (n < 11 || n > 99) ? 1 : 2",
	"(n % 10 == 1) ? 0 : ((n % 10 == 2) ? 1 : ((n % 100 == 0 || n % 100 == 20 || n % 100 == 40 || n % 100 == 60 || n % 100 == 80) ? 2 : 3))": "n==0 ? 0 : (n % 10 == 1) ? 1 : ((n % 10 == 2) ? 2 : ((n % 100 == 0 || n % 100 == 20 || n % 100 == 40 || n % 100 == 60 || n % 100 == 80) ? 3 : 4))",
	"(n % 10 == 1 && n % 100 != 11 && n % 100 != 71 && n % 100 != 91) ? 0 : ((n % 10 == 2 && n % 100 != 12 && n % 100 != 72 && n % 100 != 92) ? 1 : ((((n % 10 == 3 || n % 10 == 4) || n % 10 == 9) && (n % 100 < 10 || n % 100 > 19) && (n % 100 < 70 || n % 100 > 79) && (n % 100 < 90 || n % 100 > 99)) ? 2 : ((n != 0 && n % 1000000 == 0) ? 3 : 4)))": "n==0 ? 0 : (n % 10 == 1 && n % 100 != 11 && n % 100 != 71 && n % 100 != 91) ? 1 : ((n % 10 == 2 && n % 100 != 12 && n % 100 != 72 && n % 100 != 92) ? 2 : ((((n % 10 == 3 || n % 10 == 4) || n % 10 == 9) && (n % 100 < 10 || n % 100 > 19) && (n % 100 < 70 || n % 100 > 79) && (n % 100 < 90 || n % 100 > 99)) ? 3 : ((n != 0 && n % 1000000 == 0) ? 4 : 5)))",
	"n%10==1 && n%100!=11 ? 0 : n != 0 ? 1 : 2": "n==0 ? 0 : n%10==1 && n%100!=11 ? 1 : 2",
	"(n == 1) ? 0 : ((n != 0 && n % 1000000 == 0) ? 1 : 2)": "(n == 0) ? 0 : (n == 1) ? 1 : ((n != 0 && n % 1000000 == 0) ? 2 : 3)",
	"(n == 0 || n == 1) ? 0 : ((n != 0 && n % 1000000 == 0) ? 1 : 2)": "(n == 0) ? 0 : (n == 1) ? 1 : ((n != 0 && n % 1000000 == 0) ? 2 : 3)",
	"(n == 1) ? 0 : ((n == 2) ? 1 : ((n == 0 || n % 100 >= 3 && n % 100 <= 10) ? 2 : ((n % 100 >= 11 && n % 100 <= 19) ? 3 : 4)))": "(n == 0) ? 0 : (n == 1) ? 1 : ((n == 2) ? 2 : ((n == 0 || n % 100 >= 3 && n % 100 <= 10) ? 3 : ((n % 100 >= 11 && n % 100 <= 19) ? 4 : 5)))",
	}

I probably generated these wrongly back then.

[dwaynebailey]

These are only examples shown in Weblate, it should not affect anything else. But you're right that most of (if not all) the formulas have messed up ordering:

That is a relief. I'll see what I can do to produce a PR, if not for all of them at least for the simple common forms

[nijel]

It is likely that all of them are messed up, I did somehow generated them and apparently that was broken.

Looking at it, broken might be only the ones which don't have explicit plural numbers in the original formula. The problem with the conversion is that it turned n > 1 into n > 1 ? 0 : 1 but it actually means n > 1 ? 1 : 0. Then the n==0 was added and number increased.

[github-actions]

Thank you for your report; the issue you have reported has just been fixed.

• In case you see a problem with the fix, please comment on this issue.
• In case you see a similar problem, please open a separate issue.
• If you are happy with the outcome, don’t hesitate to support Weblate by making a donation.